Studies on the Synthesis of Phytochrome and Related Tetrapyrroles

Products 24 - 32 - found in blue-green, eucaryotic and cryptomonad algae and serve as light-harvesting proteins in photosynthesis. Phytochrome (1) pla...
0 downloads 0 Views 542KB Size
2894

J. Org. Chem. 1997, 62, 2894-2906

Studies on the Synthesis of Phytochrome and Related Tetrapyrroles. Dihydropyrromethenones by Photochemical Rearrangement of N-Pyrrolo Enamides Peter A. Jacobi,* Subhas C. Buddhu, Douglas Fry, and S. Rajeswari Hall-Atwater Laboratories, Wesleyan University, Middletown, Connecticut 06459-0180 Received February 14, 1997X

Dihydropyrromethenone 67b, a potential precursor for the synthesis of phytochrome 1, has been prepared in enantiomerically pure form beginning with N-aminopyrrole 64 and the acetylenic acid 62b. The key step involved a 3,5-sigmatropic rearrangement of N-pyrrolo enamide 66b. Introduction The biliproteins are a family of naturally occurring chromophores that are made up of linear tetrapyrrole derivatives covalently bonded to a protein (P).1-7 Representative examples include phytochrome 1, which functions as the “on-off” switch for photomorphogenesis in higher plants,6,7 the phycocyanins 2, and phycoerythrins 3 (Figure 1).8,9 Tetrapyrroles 2 and 3 are commonly

It is now well established that 1 can exist in either of two possible forms in plants: an inactive red-absorbing form known as Pr (λmax 660 nm) and an active, far red absorbing form designated as Pfr (λmax 730 nm).10-14 These two species are readily interconverted upon irradiation at 660 and 730 nm, respectively, a photoreversible-photochromic behavior which has been the subject of intensive study for many years.2-7 However, at present only the structure of the Pr form of 1 is known with some degree of certainty (Figure 2).11a,12 In the

Figure 1.

found in blue-green, eucaryotic and cryptomonad algae and serve as light-harvesting proteins in photosynthesis. Phytochrome (1) plays an essential role in many lightdependent, irreversible processes, including seed germination, flowering, and stem growth. It has also been implicated in such reversible phenomena as chloroplast movement, root tip adhesion, potassium uptake, and regulation of transmembrane potentials.2,6 Abstract published in Advance ACS Abstracts, April 15, 1997. (1) Moses, P. B.; Chua, N.-H. Sci. Am. 1988, 258, 88. (2) Phytochrome and Photoregulation in Plants, Furuya, M., Ed.; Academic Press: New York, 1987. (3) Falk, H. The Chemistry of Linear Oligopyrroles and Bile Pigments; Springer-Verlag: Vienna-New York, 1989. (4) For a review on biliproteins see: Scheer, H. Angew. Chem. 1981, 93, 230; Angew. Chem., Int. Ed. Engl. 1981, 20, 241. (5) For a review of linear tetrapyrrole chemistry see: Tetrahedron 1983, 39, 1839, Symposia-In-Print, Bonnett, R., Ed. (6) For reviews on phytochrome-mediated responses in plants, see: (a) Statter, R. L.; Galston, A. W. in Chemistry and Biochemistry of Plant Pigments; Goodwin, T. W., Ed.; Academic Press: New York, 1976; Vol. 1, p 680. (b) Ru¨diger, W.; Thu¨mmler, F. Angew. Chem., Int. Ed. Engl. 1991, 30, 1216. (c) Ru¨diger, W. Photochem. Photobiol. 1992, 56, 803. (d) Song, P.-S. The Spectrum (Bowling Green State University) 1994, 7, 1 (Issue 2). (7) For reviews on phytochrome biochemistry and chromophore chemistry, see: (a) Terry, M. J.; Wahleithner, J. A.; Lagarias, J. C. Arch. Biochem. Biophys. 1993, 306, 1. (b) Pratt, L. H. Photochem. Photobiol. 1978, 27, 81. (c) Kendrick, R. E.; Spruit, C. J. P. Photoche. Photobiol. 1977, 26, 201. (8) Schoenleber, R. W.; Kim, Y.; Rapoport, H. J. Am. Chem. Soc. 1984, 106, 2645 and references cited therein. (9) (a) Glazer, A. N. in The Biochemistry of Plants; Hatch, M. D.; Boardman, N. K., Eds.; Academic Press: New York, 1981; Vol. 8, p 51. (b) Carra, P. O.; O hEocha, C., in ref 6a, p 328. (c) Schoenleber, R. W.; Leung, S.-L.; Lundell, D. J.; Glazer, A. N.; Rapoport, H. J. Am. Chem. Soc. 1983, 105, 4072. X

S0022-3263(97)00288-0 CCC: $14.00

Figure 2.

native state Pr most likely adopts a helical geometry (all Z configuration), incorporating a 15-anti conformation.12c,d Among other theories, it has been suggested that Pfr might be derived from Pr by (a) formation of an imino ester linkage at C1, thereby extending the effective chromophore conjugation,11a,b (b) photoreversible Z,E isomerization about the C4-C5 double bond,12a and, as illustrated, (c) photoisomerization about the C15-C16 bond, with retention of a “semi-extended” chromophore conformation.12b-d According to this last model, photoisomerization induces a change in the tertiary structure of the surrounding protein shell (curves in Figure 2), thereby providing (10) Ru¨diger, W. Struct. Bonding 1980, 40, 101. See also refs 4 and 6. (11) (a) Lagarias, J. C.; Rapoport, H. J. Am. Chem. Soc. 1980, 102, 4821. (b) Micura, R.; Grubmayr, K. Bioorg. Med. Chem. Lett. 1994, 4, 2517. (12) (a) Thu¨mmler, F.; Ru¨diger, W. Tetrahedron 1983, 39, 1943. (b) Ru¨diger, W.; Thu¨mmler, F.; Cmiel, E.; Schneider, S. Proc. Natl. Acad. Sci. 1983, 80, 6244. (c) Farrens, D. L.; Holt, R. E.; Rospendowski, B. N.; Song, P.-S.; Cotton, T. M. J. Am. Chem. Soc. 1989, 111, 9162. (d) Fodor, S. P. A.; Lagarias, J. C.; Mathies, R. A. Biochemistry 1990, 29, 11141. (e) Fodor, S. P. A.; Lagarias, J. C.; Mathies, R. A. Photochem. Photobiol. 1988, 48, 129. (13) (a) Grombein, S.; Ru¨diger, W.; Zimmermann, H. Hoppe-Seyler’s Z. Physiol. Chem. 1975, 356, 1709. (b) Klein, G.; Grombein, S.; Ru¨diger, W. Ibid. 1977, 358, 1077. (c) Ru¨diger, W.; Brandlmeier, T.; Blos, I.; Gossauer, A.; Welle, J.-P. Z. Naturforsch., Teil C 1980, 35, 763. (14) (a) Scheer, H.; Krauss, C. Photochem. Photobiol. 1977, 25, 311. (b) Scheer, H.; Linsenmeier, U.; Krauss, C. Hoppe-Seyler’s Z. Physiol. Chem. 1977, 358, 185. (c) Krauss, C.; Bubenzer, C.; Scheer, H. Photochem. Photobiol. 1979, 29, 473.

© 1997 American Chemical Society

Synthesis of Phytochrome and Related Tetrapyrroles

a molecular basis for transduction of the light signal to the cells genetic regulatory apparatus. This proposal gains support from both NMR12b and SE resonance raman scattering spectroscopy (SERRS),12c,d although at present the data are not conclusive. In part this is due to the extremely small quantities of phytochrome 1 available for study from natural sources. Even in seedlings grown in the dark (etiolated), and therefore free of chlorophyll, the deep blue color of photoreceptor 1 is difficult to detect. In this paper, and the accompanying article, we describe synthetic studies which provide a basis for the preparation of naturally occurring chromophores of type 1-3 with unequivocal control over both relative and absolute stereochemistry. Ultimately these studies might lead to a better understanding of the phenomenon of photomorphogenesis.

J. Org. Chem., Vol. 62, No. 9, 1997 2895

thenone derivatives themselves, which are typically derived by coupling of monocyclic building blocks of type 4 and 5. These last two species present significant synthetic challenges in their own right, which are compounded by the fact that coupling of 4 and 5 to afford 6 is often not a trivial problem.16a,b As an alternative strategy, we were interested in the possibility that dihydropyrromethenones of general structure 6 might be prepared beginning with N-aminopyrroles of type 9 (Scheme 2). By way of summary, N-acyScheme 2

Discussion and Results Most of the published work in this area has been carried out on simple model compounds and has utilized either of two synthetic strategies. The first of these is based on biosynthetic theory and involves the oxidative cleavage of porphyrins, chlorins, and related materials.15 Although this approach can be of occasional utility when applied to unsymmetrical derivatives, it cannot provide the variety of biliproteins required for detailed study. As a more general strategy, suitably functionalized pyrromethenone derivatives of type 6 and 7 can frequently be coupled to yield linear tetrapyrrole derivatives 8 in moderate to good yields (Scheme 1).16 In principle, this Scheme 1

lation of 9 with acetylenic acid derivatives of type 10 was expected to yield the N-pyrroloamides 11, which upon 5-exo-dig cyclization would give N-pyrrolo enamides of general structure 12. Enamides 12, upon 3,5-sigmatropic rearrangement17 and subsequent aromatization, would then afford dihydropyrromethenones 6 with complete control over both relative and absolute stereochemistry. An attractive feature of this strategy was the fact that stereochemical and regiochemical features incorporated into 10 would be transposed in an unequivocal fashion to the final product 6. As will be reported, there was reason to believe that acyclic intermediates 10 could be synthesized in enantiomerically pure form using a Nicholas reaction (dashed line in 10, vide infra).18 The feasibility of this strategy was initially tested with the simple model systems 17 and 18, which, because of their symmetrical nature (C, D ) H, cyclohexyl), were readily prepared by following standard literature proceScheme 3

approach provides satisfactory control over both stereoand regiochemical features (A-H in 8) as well as oxidation state at crucial ring positions. However, this second strategy is limited by the availability of the pyrrome(15) See, for example: (a) Smith, K. M.; Kishore, D. Tetrahedron 1983, 39, 1841. (b) Cavaleiro, J. A. S.; Smith, K. M. J. Chem. Soc., Perkin Trans. 1 1973, 2149. (c) Barnett, G. H.; Hudson, M. F.; McCombie, S. W.; Smith, K. M. J. Chem. Soc., Perkin Trans. 1 1973, 691. (d) Smith, K. M.; Sharkus, L. C.; Dallas, J. L. Biochem. Biophys. Res. Commun. 1980, 97, 1370. (e) Bonnett, R.; McDonagh, A. F. J. Chem. Soc., Perkin Trans. 1 1973, 881. (16) See, for example (a) Bishop, J. E.; O’Connell, J. F.; Rapoport, H. J. Org. Chem. 1991, 56, 5079. (b) Bishop, J. E.; Nagy, J. O.; O’Connell, J. F.; Rapoport, H. J. Am. Chem. Soc. 1991, 113, 8024. (c) Bishop, J. E.; Dagam, S. A.; Rapoport, H. J. Org. Chem. 1989, 54, 1876. (d) Schoenleber, R. W.; Kim, Y.; Rapoport, H. J. Am. Chem. Soc. 1984, 106, 2645 and references cited therein. (e) Gossauer, A.; Hirsch, W. Liebigs Ann. Chem. 1974, 1496. (f) Gossauer, A.; Hinze, R.-P. J. Org. Chem. 1978, 43, 283. (g) Gossauer, A.; Weller, J.-P. Chem. Ber. 1980, 113, 1603.

(17) Preliminary communications: (a) Jacobi, P. A.; Buddhu, S. C. Tetrahedron Lett. 1988, 29, 4823. (b) Jacobi, P. A.; Rajeswari, S. Tetrahedron Lett. 1992, 33, 6231. See also: (c) Patterson, J. M.; Ferry, J. D.; Boyd, M. R. J. Am. Chem. Soc. 1973, 95, 4356. (18) (a) Schreiber, S. L.; Klimas, M. T.; Sammakia, T. J. Am. Chem. Soc. 1987, 109, 5749. (b) Schreiber, S. L.; Sammakia, T.; Crowe, W. E. J. Am. Chem. Soc. 1986, 108, 3128. See also: (c) Lockwood, R. F.; Nicholas, K. M. Tetrahedron Lett. 1977, 18, 4163. (d) Nicholas, K. M.; Nestle, M. O.; Deyferth, D. Transition Metal Organometallics; Alper, H., Ed.; Academic Press: New York, 1978; Vol. 2, p 1. (e) Evans, D. A.; Ennir, M. D., Mathre, D. J. J. Am. Chem. Soc. 1982, 104, 1737.

2896 J. Org. Chem., Vol. 62, No. 9, 1997

dures (Scheme 3).19 Thus, condensation of N-aminophthalimide (14) with dialdehydes 13a,b gave an excellent yield of the protected N-aminopyrroles 15a,b, which could be directly cleaved to the aminopyrroles 17a,b with hydrazine in ethanol or converted to the methyl esters 16a,b with oxaloyl chloride/AlCl3 followed by methanolysis.20 Hydrazinolysis of 16a,b then proceeded routinely to afford the amino esters 18a,b with no complications due to ester aminolysis. Once in hand, both 17a,b and 18a,b were cleanly coupled with the acetylenic acid 19 to provide the hydrazide derivatives 20a,b and 21a,b (Scheme 4). As expected for electron deficient alkynes, Scheme 4

Jacobi et al.

26a, and 28a, together with a larger proportion of the cleavage products 30 and 31a. Similar results were obtained at 253 nm. Significantly, cleavage products 3032 were the only products observed in the presence of triplet sensitizers. In the presence of piperylene (triplet quencher),21 however, cleavage was reduced to trace amounts, and 24a was obtained in 40-50% yield as an equilibrium mixture of E and Z isomers (∼1:1). Similar results were obtained with 22b, and in identical fashion, 23a,b afforded 40-50% yields of the target pyrromethenones 25a,b.22 These studies are consistent with a reaction pathway in which photodissociation occurs via a triplet state, in competition with a singlet state 3,5sigmatropic shift. Although the yields obtained in the conversion of 22 and 23 to the dihydropyrromethenones 24 and 25 were not as high as might be desired, we were sufficiently encouraged to pursue additional studies with substrates bearing the natural substitution pattern. In order for these preliminary studies to be extrapolated to the preparation of dihydropyrromethenones of general structure 35 (a logical precursor to 1-3), it was first necessary to devise efficient syntheses of both N-aminopyrroles of type 33 and highly substituted acetylenic acids of type 34 (Figure 3). As in the case with

these last materials then underwent a facile 5-exo-dig cyclization to afford either 22a,b or 23a,b in >90% yield (∼3:1 mixture of E- and Z-isomers). This step completed the formation of rings A and B. Numerous conditions were examined for converting 22a,b and 23a,b to the isomeric pyrromethenones 24a,b and 25a,b (Scheme 5). These materials were stable to Figure 3.

Scheme 5

N-unsubstituted pyrroles, the synthesis of 33 required strict control of regiochemistry, which turned out to present a significant challenge. Ultimately, however, these materials were derived by following the route outlined in Scheme 6, which takes advantage of a highly ortho-selective Diels-Alder reaction of 2-alkoxy-1,3-pentadiene derivatives 36 with 2-oxo-3-butenoate esters 37.23 Scheme 6

thermolysis at temperatures up to 250 °C,17c and at higher temperatures they suffered only slow decomposition to intractable tars. Also, all attempts at acid catalysis led to decomposition. Upon photolysis, however, 22a,b and 23a,b gave reaction mixtures which contained trace amounts of the desired products of 3,5-sigmatropic shift (24, 25), in addition to products corresponding to 1,3- and 1,5-sigmatropic shifts (26-29) and N-N bond cleavage (30-32). After considerable experimentation, we found that the ratio of products 24-32 was strongly influenced by the presence or absence of triplet state quenchers. For example, at 300 nm 22a (E- or Z-isomer) gave 5-10% yields of the rearrangement products 24a, (19) Zimmerman, H.; Flitsch, W.; Kramer, V. Chem. Ber. 1969, 102, 3268. (20) Kakushima, M.; Hamel, P.; Frenette, R.; Rokach, J. J. Org. Chem. 1983, 48, 3214.

Adducts 38 were then converted to protected N-aminopyrroles 40 by a two-step sequence involving ozonolysis to afford 1,4-dicarbonyl species 39, followed by Paal-Knorr cyclization with N-aminophthalimide (14). Finally, as described in Scheme 3, hydrazinolysis of 40 gave a virtually quantitative yield of the target pyrroles 33. (21) (a) Hammond, G. S.; Turro, N. J.; Leermakers, P. A. J. Phys. Chem. 1962, 66, 1144. (b) Yang, N. C.; Hui, M. H.; Shold, D. M.; Turro, N. J.; Hautala, R. R.; Dawes, K.; Dalton, J. C. J. Am. Chem. Soc. 1977, 99, 3023. (22) The structure of dihydropyrromethenone 25b (E isomer) and acetylenic acid 62c were unequivocally established by single-crystal X-ray analysis: performed by Ms. Gayle Schulte, Yale University. (23) Jacobi, P. A.; Cai, G. Heterocycles 1993, 35, 1103.

Synthesis of Phytochrome and Related Tetrapyrroles

Since the ozonolysis products 39 were generally not isolated, this route constitutes a convenient three-step sequence for preparing 33 from readily available starting materials. Our synthesis of acetylenic acids 34 built upon work by Schreiber et al., who first demonstrated that Nicholas alkylations can be carried out with high enantioselectivity.18 In an elegant mechanistic study, this group observed that Bu2BOTf-catalyzed condensation of Evans’ enolate 41 with the cobalt complex 42 occurs with kinetic resolution,18e affording an ∼80% yield of the syn-adduct 43s having exclusively the S,S configuration (Scheme 7; syn:anti selectivity ) 12:1).18a Oxidative removal of cobalt Scheme 7

J. Org. Chem., Vol. 62, No. 9, 1997 2897

This result is in general accord with the observations of Schreiber et al., who noted that selectivity increases with increasing size of Y.18a,b Equally disappointing, we were unable to selectively remove the chiral auxiliary in 50 without concomitant hydrolysis of the acetylenic ester to afford diacid 51.24 This lack of differentiation was a serious complication, since all attempts at monofunctionalization of 51 invariably led to complex mixtures of products. In contrast to the poor selectivity observed with esters 49a-c (Scheme 8), trimethylsilyl derivative 42 underwent clean condensation with oxazolidinone 48, affording Nicholas adduct 53 in 90-95% yield with >98% syn selectivity (Scheme 9; this selectivity is significantly higher than that observed with oxazolidinone 4118b). This Scheme 9

then gave a quantitative yield of the alkyne 44s. This work provided excellent precedent for the synthesis of acetylenic acids of type 34 (Scheme 7). By analogy, we were confident that reaction of aldehydes of general structure 45 with acetylides 46 would afford propargyl alcohol derivatives 47, which upon Nicholas-Schreiber condensation with an appropriate chiral enolate would give ring-A precursors 34 with unequivocal control over stereochemistry at C2, C3, and C3′ (tetrapyrrole numbering). The flexibility of this approach might be put to good advantage in confirming the postulated relative and absolute stereochemistry in 1-3. We initially expected that alkynes of type 34 would be of greatest utility when Y ) carbalkoxy, since it appeared that an electron deficient triple bond was required for 5-exo-dig cyclization (cf. Scheme 4). Therefore, our preliminary studies focused on preparing simple dimethyl analogs of type 50a-c, which were synthesized in analogous fashion to 44s but employing the chiral oxazolidinone 48 (Scheme 8). In this case, however, we were disappointed to find that syn-selectivity in the reaction 48 + 49 f 50 was only ∼3:1 (60-80% yield). Scheme 8

reaction clearly demonstrated the potential for achieving stereoselectivities of the level desired for the synthesis of 1-3. Adduct 53 was then readily converted to the acetylenic hydrazide 55a (Y ) H) by a two-step sequence involving hydrolysis to the acetylenic acid 54a (concomitant removal of TMS group)24 and EDCI-catalyzed coupling with N-aminopyrrole 18b. At this stage, we experienced considerable difficulty in effecting the required 5-exo-dig cyclization leading from 55a to enamide 56a (Scheme 9). Not surprisingly, 55a was inert to cyclization under thermal conditions, and it rapidly decomposed upon attempted acid or base catalysis. These results are in marked contrast to the ease of cyclization of activated alkynes of type 20 and 21 (cf. Scheme 4). In addition, solvomercuration-demecuration took place mainly with participation of the hydrazide carbonyl group to give modest yields of cyclic imino esters. Eventually, some degree of success was achieved with the reagent system PdCl2(MeCN)2/NaOAc, which afforded 60-70% yields of the desired enamide 56a.25,26 This reaction was also accompanied by significant amounts of alkyne coupling. However, by far the most useful procedure was discovered in a serendipitous fashion upon attempted cleavage of the trimethylsilyl group from acetylenic hydrazide 55a′ (Y ) TMS). This (24) Evans, D. A.; Britton, T. C.; Ellman, J. A. Tetrahedron Lett. 1987, 28, 6141. (25) Rudisill, D. E.; Stille, J. K. J. Org. Chem. 1989, 54, 5856. (26) As expected, enamides 56 exhibited atropisomerism due to hindered N-N bond rotation, although each isomer had identical photochemical behavior. See, for example, ref 3, p 108.

2898 J. Org. Chem., Vol. 62, No. 9, 1997

material was obtained from adduct 53 by hydrolysis and amidation under carefully controlled conditions. 55a′ afforded none of the expected terminal alkyne 55a upon being warmed with n-Bu4NF (TBAF), but rather was directly converted to the identical N-pyrrolo enamide 56a obtained from Pd(II)-catalyzed cyclization of 55a. The same conditions, when applied to terminal alkyne 55a, afforded enamide 56a in 70-90% yield. The precise mechanism by which the fluoride ion catalyzes the cyclization of 55a,a′ to 56a is not known with certainty, but it presumably involves a strong hydrogen bond between F- and the hydrazide N-H group, with an attendant increase in N-nucleophilicity.27 In any event, we utilized an identical two-step sequence to convert adduct 44s to the enantiomeric hydrazide ent-55a (ent ) mirror image of structure shown), which when warmed with TBAF gave an excellent yield of enamide ent-56a. As with 56a above, ent-56a was obtained as a single enantiomer. With the problem of hydrazide cyclization apparently solved, we turned our attention next to preparing acetylenic acids having the proper constitution for eventual conversion to 1-3. In phytochrome (1) the absolute stereochemistry at C2 and C3 has been assigned as R, but it is important to maintain as much flexibility as possible in the synthetic scheme. As summarized in Scheme 7 (45 f 34), we intended that both relative and absolute stereochemistry at C2-C3 would be controlled through the use of an enantioselective Nicholas reaction (vide supra), while stereochemistry at C3′ (also believed to be R) would be established by utilizing an appropriate aldehyde 45 from the “chiral pool”. In principle, the aldehyde chosen could incorporate a sulfur ligand of proper absolute configuration from the start (X ) S-R). Alternatively, the desired configuration could be obtained by nucleophilic displacement with inversion of an activated hydroxyl group at a later stage of the synthesis (X ) O-R).28 This second approach offered a greater degree of flexibility, and it also had the advantage that both Rand S-R-hydroxyaldehyde derivatives of the required composition are readily available from (R)- and (S)-lactic acid, respectively.29 Our expectations regarding the utility of the Nicholas reaction turned out to be fully justified (Scheme 10).

Jacobi et al.

(LiTMSA) with aldehydes 57b and 57c afforded 90-95% yields of the corresponding acetylenic alcohols 58b,c,29 which without isolation were methylated (DMS) to give the methyl propargyl ethers 59b,c in excellent overall yield. Conversion of 59b,c to the cobalt complexes 60b,c was then accomplished by following standard literature procedures.18 Reaction of 60b,c with the chiral boron enolate 48 then gave the Nicholas adducts 61b,c (>95%), which upon hydrolysis provided the target (2R,3R,3'S)acetylenic acid derivatives 62b,c in 60-70% overall yield from aldehydes 57.22 In both cases syn-stereoselectivity was >98%. In identical fashion, (2S,3S,3′R)-acetylenic acids ent-62 were prepared with similar yields and selectivities by utilizing the boron enolate 41. As described above for the acetylenic acids 54 and ent54 (Scheme 9), acetylenic acids 62b,c were readily converted to the corresponding N-pyrrolo enamides 56b,c by a two-step sequence involving EDCI-mediated coupling with N-aminopyrrole 18b, followed by TBAFcatalyzed cyclization (Scheme 11). Enantiomerically pure Scheme 11

Scheme 10

enamides 56a-c and ent-56a, as well as achiral enamide 56d, were then subjected to photochemical rearrangement, using conditions similar to those employed for model systems 22 and 23 (300 nm, tert-amyl alcohol, piperylene, -10 °C; cf. Scheme 5). In general, yields for this step were moderate to good, ranging from a low of 37% for 63b to 78% for 63d (see table). Dihydropyrromethenones 63 were obtained as ∼1:1 mixtures of E and Z isomers. In analogous fashion, enamide ent-56a gave the enantiomeric dihydropyrromethenone ent-63a, which within experimental error had equal but opposite [R]25D to that observed for 63a (Z isomers).

Thus, condensation of lithium (trimethylsilyl)acetylide

(27) (a) Pless, J. J. Org. Chem. 1974, 39, 2644. (b) Clark, J. H. Chem. Rev. 1980, 80, 429. (c) Morrison, H. J. Am. Chem. Soc. 1965, 87, 932. (d) Jacobi, P. A.; Rajeswari, S. Tetrahedron Lett. 1992, 33, 6235. (28) Volante, R. P. Tetrahedron Lett. 1981, 22, 3119 and references cited therein. (29) Takai, K.; Heathcock, C. H. J. Org. Chem. 1985, 50, 3247 and references cited therein.

Synthesis of Phytochrome and Related Tetrapyrroles

As in the case with enamides 22 and 23 (Scheme 5),17a satisfactory yields of 63 and ent-63 were obtained only in the presence of piperylene (triplet quencher), which minimizes the formation of byproducts arising from hydrazide cleavage. Interestingly, benzyl ether 56c (A ) Me, B ) S-CHOBnCH3) and methyl ether 56b (A ) Me, B ) S-CHOMeCH3) showed markedly different behavior upon attempted photochemical rearrangement. Thus, 56b afforded a ∼40% yield of dihydropyrromethenone 63b after 21 h at -10 °C (300 nm), while 56c reacted only very slowly to give mainly the products of hydrazide cleavage (